It's time to generate for real. This module walks you through the easiest way: the web interface. You launch the app with one command, get to know each part of the screen, enter the company name, follow the 3 phases live, and download the finished package—pointing and clicking, without memorizing commands.
Illustrative diagram — from command to download: the web app handles the three phases and delivers the package at the end.
The web mode is the most user-friendly way to use the tool — and deploys with a single command. Right after that, the browser opens automatically on the tool’s page.
# lembre de ativar o venv antes
python -m strategy_factory.webapp
# o navegador abre em:
http://localhost:8888This is the only command you need to use the tool on the web.
While you use the interface, the terminal needs to stay open and run the command—it’s what keeps the app running. To shut it down afterward, just close the terminal or press Ctrl+C in it.
Before running it, it’s worth getting to know the screen. It is simple by design: basically a field, a selector, an optional field, and a button. Getting to know each part gives you confidence for your first analysis.
Illustrative recreation of the screen — not an actual screenshot of the app.
This is the only required field. Enter the name of the company you want to analyze — for example, "Stripe". It’s the starting point for the entire analysis.
# você digita: Stripe
# os arquivos vão para:
output/stripe/
├── markdown/ (15 documentos)
├── presentations/ (2 apresentações)
├── documents/ (2 relatórios)
└── mermaid_images/ (diagramas)The name you enter becomes the name of the folder (in lowercase) where everything is saved.
If the name could be confusing (more than one company has that name), don’t worry: the context field in the next section helps you specify exactly which company you mean.
The context field is optional, but it makes a difference. With a few details — industry, size, business model — you help Perplexity research much more accurately.
# exemplos de contexto útil
fintech, B2B, 200 funcionários
varejo de moda, e-commerce, mercado brasileiro
SaaS de RH, foco em PMEs, série AThe more specific you are, the better the result — especially for lesser-known companies.
E.g., fintech, retail, healthcare, education. This already narrows the research considerably.
E.g., B2B or B2C, number of employees, stage (startup, Series A...).
E.g., Brazilian market, Latin America. Clears up ambiguity about the name.
For well-known brands, the name alone is enough. But for local, new, or commonly named companies, context is what separates an accurate report from a generic one.
After starting, a bar shows progress through the three phases in real time. You can see exactly what’s happening and tell as soon as the package is ready.
The bar shows the collection of information about the company and its industry.
You watch the documents being generated one by one.
The final step builds presentations, Word documents, and diagrams on your computer.
The phases run in sequence: Synthesis starts only when Research is finished, and Generation starts only at the end. That’s why the bar moves in stages—it’s normal for each phase to take a little while before moving on to the next.
This is the end goal: have the files on hand. When it finishes, the screen shows download links for each deliverable — and everything is also saved on your machine.
Each deliverable (documents, presentations, Word files, and diagrams) has its own link to download it directly from the interface.
Everything is also saved in output/company-name/, even if you close the browser.
You can view the contents of the markdown documents directly in the interface before downloading them.
15 documents + 2 presentations + 2 reports + diagrams, ready to review and present.
output/stripe/
├── markdown/ (15 .md — os documentos de texto)
├── presentations/ (2 .pptx — resumo e completa)
├── documents/ (2 .docx — relatório e proposta)
└── mermaid_images/ (.png — os diagramas)Even if you don’t download them from the page, all the files are here.
With this, you’ve completed the full cycle: understand, install, and generate. In Track 2, you’ll use the tool in depth (command line, modes, and settings); in Track 3, you’ll explore what to do with each of the 15 documents.
Track 2 — In-Depth Use — master the command line, quick and in-depth modes, context usage, and how to resume interrupted analyses.